DE68914898T2 - Selective removal of chemical species from a fluid. - Google Patents

Description

The invention relates to the selective removal of a chemical species from a fluid.

It is known that chemical species in a fluid, e.g. an aqueous solution or suspension, can be selectively removed and thereby separated from another species by contacting the fluid with a solid material having centers capable of reacting selectively with the species to be removed. Various forms of reaction are possible. For example, the reaction may be predominantly chemical, as in the case of ion exchange or chelation, or it may be biochemical in nature, as in the case of the formation of affinity complexes between biochemical molecules, cells and other particles. If the reaction is reversible, the species that specifically reacted can be recovered.

The centers on the solid material at which a reaction occurs may comprise an atom, a group of atoms, or a molecule, usually referred to as a ligand, that is bound to the material and is capable of reacting selectively with the species to be removed.

According to one aspect, the invention makes use of the application of biospecific affinity reactions and in particular of immunochemical reactions. In this case, the ligand can be an immunochemical component, e.g. a protein that has a biospecific affinity for another immunochemical component. An important class of such reactions is that between an antigen (or hapten) and an antibody directed against it. The use of monoclonal antibodies enables highly specific affinity reactions to be carried out.

Examples of antigens, often referred to as foreign substances, include viruses, bacteria, bacterial toxins, hydrocarbons, hormones, drugs and lectins.

In addition, other reactions between biologically active compounds, such as enzymes and their substrates, are often sufficiently specific and have sufficient affinity to be used for separations. In addition, small molecules, such as amino acids, can react specifically with and with sufficient affinity to, for example, proteins, so that separation can be achieved.

A known method for removing a chemical species from a solution is based on the use of an affinity chromatography column. This column is filled with particles of a porous polymer gel, e.g. cross-linked agarose, to which a ligand is covalently bound. For example, such polymer gels are described in US-A-4,330,440. From this patent and other prior art it is known that a ligand can be covalently bound to an activated polymer, i.e. to a polymer with active groups present on its surface with which a ligand can react. Suitable active groups include functional groups which, if necessary, can be generated in an activatable polymer by treating the polymer with an activating agent. For example, an activatable polymer such as agarose with hydroxyl groups on its surface can be treated with a carbonylating agent or cyanogen bromide to create functional groups on its surface that can react with a protein ligand.

When using the affinity chromatography column, solution containing the species to be removed is passed through the column and between the particles so that the species is bound to the ligands. This method has a number of disadvantages. These include the fact that the polymer gel is difficult to handle. Furthermore, since the particles are compressible, the packing of the beads in a column can lead to excessive back pressure when the solution is passed through the column. Furthermore, the diffusion of the solution through the pores of the polymer particles is a relatively slow process and the particles can act as a filter for particulate material such as cells. Another disadvantage is that the ligand is distributed throughout the porous structure of the particles, with varying degrees of accessibility and, as a result, is not effectively utilized.

In another known method, the solution containing the chemical species to be removed is passed through a filter having a fibrous matrix, e.g. filter paper or a porous polymeric membrane, to which the ligand is bound. Such filters suffer from the disadvantage that they may have a tendency to block when particulate material is passed through the filter and may also have the disadvantages associated with pore diffusion and ligand usage as described above.

US-A-4 551 435 describes a method for the selective removal of immunospecifically recognizable substances from biological fluids such as blood and bone marrow. Undesirable antigens or antibodies can be filtered out of the blood by first binding the antigen or antibody with the corresponding complementary antibody or antigen complex. Immune complexes formed in this way are then passed through an adsorbent material that has an affinity for immune complexes. A ligand consisting of a non-immunospecific factor, e.g. Clq, is coupled to a solid phase to form the adsorbent.

The affinity separation described in US-A-4 551 435 can be carried out by passing the fluid through a chamber comprising a spiral wound sheet with opposing surfaces for receiving the non-immunospecific factor. The spiral wound Convolutions of the sheet material are separated from each other by spacers, and when the chamber is in use, fluid flows through the convolutions, either axially or peripherally. The only specific example of an adsorbent given is a polystyrene sheet to which the non-immunospecific factor Clq is bound.

A major disadvantage associated with the use of a polystyrene sheet to which the factor Clq is bound is that the factor Clq is adsorbed and not covalently bound to the sheet and there is a risk of desorbing. Furthermore, the polystyrene surface can adsorb other biologically active compounds from the liquid that are passed over the surface, which compounds can act as additional ligands. Moreover, when adsorption occurs, there is no control over the orientation of the adsorbed species, with the result that the activity and/or specificity of the ligand can be compromised. In this regard, the ligand that can be used in the case of the method and immunoadsorbent material according to US-A-4,551,435 is limited to a non-immunospecific factor. Another disadvantage associated with the use of a polystyrene sheet is that it has a hydrophobic surface that can inactivate a protein.

The aim of the present invention is to solve the above-described problems that occur in the case of the prior art.

The present invention provides an element for the selective removal of a chemical species from a fluid, comprising an impermeable carrier sheet, the surface of which is designed to receive a ligand for the species to be removed, the element being characterized in that a layer made of a polymer which has functional groups on its surface through which the ligand is directly or indirectly covalently bound to the polymer.

The invention further provides a method for selectively removing a chemical species from a fluid comprising contacting the fluid with an element according to the invention.

According to another aspect of the invention there is provided an apparatus for the selective removal of a chemical species from a fluid comprising a housing defining a chamber, the housing having fluid inlet and outlet means, the chamber containing at least one element having a ligand for the species to be removed and the element or elements being arranged relative to the inlet and outlet so as to form a flow path such that, when the apparatus is in use, fluid entering the chamber through the inlet is passed over the surface of the element or elements before the fluid leaves the chamber again through the outlet, the apparatus being characterized in that the element is an element according to the invention.

Preferably, the flow path is such that the depth of the fluid passed over the element or elements is 20-500 µm, more particularly 30-200 µm.

The invention will be illustrated by way of example in the attached schematic drawings, in which the drawings show:

Fig. 1 is a cross-sectional view of a preferred element of the invention;

Fig. 2 is a perspective view of the exterior of a device according to the invention;

Fig. 3 is a longitudinal sectional view taken along lines 2-2 of Figure 2; and,

Fig. 4 is a cross-sectional view taken along lines 3-3 of Fig. 2.

The support sheet of the element of the invention can be formed from a variety of different materials. A suitable material can be, for example, a metal, glass or a polymer. Many polymeric materials that can be formed into a sheet or film are suitable, including, for example, cellulose ethers, or esters, e.g., cellulose acetate, polyesters, e.g., poly (ethylene terephthalate), polyolefins, e.g., poly (propylene) and poly (vinyl chloride).

The thickness of the support can vary greatly depending on the material from which it is made and the way in which the element is used. For reasons of compactness, the support sheet is preferably as thin as possible, while still meeting the mechanical stability requirements. For example, the thickness of the support sheet can be 0.01-0.5 mm, particularly advantageously 0.05-0.2 mm.

Preferably, the carrier sheet is flexible. Preferably the carrier sheet is also flat.

The polymer layer can be in the form of an activated polymer layer, i.e. with functional groups that react directly with the ligand. Alternatively, it can be in the form of an activatable polymer layer that is subsequently activated by treatment with an activating agent. The activating agent can convert a functional group of the activatable polymer into a functional group that is capable of reacting with the ligand or the activating agent can be a coupling agent that is bound to the polymer by reaction with a functional group of the activatable polymer.

Examples of activating agents include divinylsulphone, cyanogen bromide and glutaraldehyde.

Suitable polymers can be derived from monomers such as ethylenically unsaturated hydroxy group-containing monomers, e.g. hydroxyethyl methacrylate (HEMA), ethylenically unsaturated oxirane group-containing monomers, e.g. glycidyl acrylate and ethylenically unsaturated amide group-containing monomers, e.g. acrylamide.

To achieve the removal of a selected chemical species from a fluid, e.g. the removal of a specific protein from a mixture thereof, the polymer is preferably designed such that non-specific adsorption to the element is minimal.

Preferably, the activated or activatable polymer is practically hydrophilic. Particularly suitable chemical groups that contribute to the hydrophilicity of the polymer include hydroxyl, amino, carboxyl and thiol groups.

In order to minimize the problems associated with the use of porous materials, the activated or activatable polymer layer can be practically non-porous. If the activated or activatable polymer layer is porous, the pores are preferably sufficiently small to prevent the ligand or the species to be removed, e.g. a protein, from entering the layer. Preferably, the activated polymer layer is practically non-swellable.

For ease of manufacturability, the activatable or activated polymer is preferably coatable from a solvent, e.g. from a solution in water and/or an organic solvent. In this way, conventional coating equipment can be used, including, e.g., slide coating hoppers or extrusion coating hoppers, to efficiently produce large quantities of coated product.

Preferably, the activatable or activated polymer layer forms a continuous layer on the carrier.

The thickness of the activated polymer layer depends on such factors as the particular polymer used. Since in a preferred embodiment the reaction between the ligand and the species to be removed takes place predominantly on the surface of the layer, it need only be sufficiently thick to effect adequate binding of the ligand to the carrier sheet. For example, the dry density of the activated polymer layer may be 5-100 µm, most advantageously 10-50 µm.

Adequate adhesion between the activated polymer layer and the carrier sheet can be achieved by appropriate selection of the two materials used. Alternatively, adhesion can be enhanced by other measures, for example by using an adhesion-enhancing layer or by subjecting the carrier sheet to a corona discharge or an RF plasma treatment before applying the polymer layer.

In a preferred embodiment of the invention, a layer of activated polymer is provided on each side of the support sheet.

Various methods can be used to maximize the surface area of the activated or activatable polymer layer relative to the surface area of the carrier sheet. In one such method, the layer contains an intrinsic particulate material that increases the surface area of the layer in the vicinity of each particle. The particulate material can be in the form of beads. Suitable materials from which the particles can be composed include polymers and glass.

According to a particularly preferred embodiment of the invention, the activated or activatable polymer layer contains a particulate material which serves as a spacer, ie the particulate material forms the means whereby an element of the invention can be spaced from another element or from another part of the same element. The particulate material can be chemically held on or within the polymer coating. Preferably, the particulate material comprises particles of substantially uniform shape and dimension. In many applications, it is desirable that the individual particles be distributed within the layer such that a substantially uniform distance of separation between adjacent elements is achieved.

As described above, the particulate material may have a variety of shapes, including, for example, the shape of beads of a polymer or glass. The dimension of the particles, which determines the degree of spacing, depends on such factors as the separation distance required between adjacent elements and the thickness of the activated polymer layer. Substantially spherical beads of an inert material, all of substantially the same diameter within a range of 20-500 µm, preferably 30-200 µm, represent an example of a suitable particulate material.

A particularly advantageous feature of the particulate spacing means described above is that it is possible to apply the particulate material to the support with the polymer layer. In this way, an element with integral spacing means can be produced. By simply preparing a homogeneous coating composition with the polymer or monomers forming the polymer layer and the particulate material, the particles can be uniformly distributed over the applied layer, thereby enabling uniform separation.

The use of beads or other suitable particulate material adhered to the surface of a substrate as a spacing agent is considered novel.

Generally, before the element can be used to remove a substance from a fluid, e.g., an aqueous solution or suspension, a ligand for the substrate is covalently bound to the activated polymer layer of the element. In a preferred embodiment of the invention, where the polymer layer is substantially nonporous, the ligand is bound predominantly to the surface of the layer.

The process for removing a substance from a fluid may be carried out in the device of the invention. The element of the invention contained in the device may be designed in a number of different ways.

For example, the device may comprise a plurality of elements facing each other, each element being separated from adjacent or adjacent elements by spacing means.

According to a preferred embodiment of the device, the element is in the form of a coil, the turns of the coil being separated from each other by spacers and the distinct current path runs axially relative to the axis of the coil.

In another preferred embodiment, the element is in the form of a coil, the turns of the coil being separated from one another by spacers and the defined current path running around the circumference of the turns.

Preferably, the spacing means provides a substantially constant separation distance between adjacent surfaces of the element or elements. The separation distance may be 20-500 µm, preferably 30-200 µm.

As described above, the spacing means may be an integral part of the element. Alternatively, the spacing means may be separate means and may be in the form of, for example, a band, rods or a mesh-like structure which allows the flow of fluid through the device.

Where it is not important that the spacing means provide a substantially uniform separation distance, the spacing means may be provided by the element itself. For example, the element may be corrugated or ribbed and adjacent parts of the element or adjacent separate elements may be arranged so that only parts of the element or elements abut one another.

According to one aspect of the invention, the element or device can be used to remove a biologically active molecule from a mixture thereof by using a biospecific affinity reaction. In this case, the ligand may be a biospecific ligand, e.g. a protein, and the activated polymer will have suitable functional groups by which the protein can be covalently bound to the polymer. Such groups are well known and include hydroxyl, amine and thiol groups. An example of a protein ligand is Protein A.

Many biospecific affinity reactions are known and a particular example is based on the reaction of an antigen and an antibody directed against it. In view of the wide range of monoclonal antibodies available, it is now possible to carry out a wide range of highly specific affinity reactions. A particular antibody ligand that has been used in the invention is a rat anti-mouse kappa chain antibody

Other ligands that can be used include reactive dyes such as triazine dyes, e.g. Procion MX-R and Cibacron Blue F3G-A.

The advantages of using the device of the invention include the fact that it is suitable for processing fluids containing particulate materials, e.g. cells, and the device is much less susceptible to blockage by such particulate material compared to available devices. Furthermore, the device is self-contained and disposable, making it suitable for single use if desired. This is of great importance when handling materials containing substances such as pathogens, viruses or DNA products, or when the treated fluid must be re-injected into a patient (e.g. in bone marrow purge). In addition, the device can be easily pre-packaged and pre-sterilised if desired.

It will be appreciated that the device may be supplied in various forms taking into account the nature of the element contained in the device. For example, the polymer layer may be in an activatable form, so that before using the device the element must undergo treatment with an activating agent and subsequently treatment to bind the ligand. Alternatively, the polymer layer may already be activated and only require treatment to bind the ligand. Finally, the device may be supplied with a ligand bound to the element.

The element and device of the invention are further described with reference to Figs. 1-4 (not to scale).

Fig. 1 shows a cross-sectional view of an element according to the invention. The element comprises a carrier sheet 10 coated with a layer of an activatable or activated polymer 11. Beads 12 present in the layer 11 adhere to the carrier 10.

Fig. 2 is a perspective view of the exterior of a Device according to the invention. The housing 20 is shown, which can be formed from a plastic material, eg polypropylene. The housing 20 comprises a cylindrical body part 21 which is provided with a cap 22. The cap is provided with a fluid inlet tube 23 and the body part is provided with a fluid outlet tube 24.

Figures 3 and 4 are cross-sectional views taken along lines 2-2 and 3-3 of Figure 2, respectively.

The cap 22 contains an axial passage 25 through which fluid can be introduced into the chamber defined by the housing 20. The inner surface of the cap 22 is provided with grooves 26 extending radially from the passage 25 to spread the fluid flow as it enters the chamber. The chamber contains a spool or reel 27 of an element according to the invention. An element of the type shown in Fig.1 is spirally wound around a cylindrical core 28. (The polymer layer 11 is not shown in Figs. 3 and 4). The outer turn of the reel or reel is attached to the body of the reel or reel by means of an adhesive tape 29 which is co-extensive with the outer surface of the spool or reel and provides a liquid-tight seal between the spool or reel and the inner surface of the housing 20.

The reel or spool fills the chamber between the cap 22 at one end and a polypropylene disk 30 which presses against the annular wall of the chamber at the other end. The surface of the disk facing the reel or spool is provided with grooves 31 which extend radially from a central passage and pass axially through the disk. This passage communicates with the passage 32 which leads through the wall of the housing and the outlet pipe 24.

In use of the device, fluid entering the chamber through the inlet passes through the turns of the reel or spool before leaving the chamber through the outlet.

It should be noted that the drawings, and in particular the representation of the coil or reel, are schematic. In practice, the total thickness of the element may be of the order of 200 µm. A reel of such an element was made in the form of a strip 35 mm wide and 11 m long, wound spirally around a central cylindrical core having a diameter of about 12 mm. Such a coil was contained in a device of the type shown, with a total length of 80 mm and an external diameter of 70 mm. Of course, the coil or reel consists of many closely spaced turns which cannot be represented in a scaled drawing.

The invention is further described by means of examples as follows.

example 1

A coating solution containing 10% w/w polyhydroxyethyl methacrylate (HEMA) in ethanol and glutaraldehyde as crosslinking agent in an amount of 5% w/w polymer was prepared.

An extrusion hopper was used to apply the solution to one side of a corona discharged polyethylene terephthalate sheet of 0.08 mm thickness. The solution flow rate was adjusted to achieve a wet coating of 100 µm at a coating speed of 3 m/min. The layer was dried at room temperature in the coater for about 20 min. and cured at 90ºC for about 2 days.

The coating operation was also carried out on the other side of the sheet, with the modification, however, that the solution contained silica-coated polystyrene spacer beads of 50 µm diameter to provide a density of about 200 beads/cm2 in the dried product.

The dry density of the cross-linked polymer layer was 10-20 μm and had minimal swelling in water.

The sheet produced in this way was cut into strips 11 m long and 35 mm wide each.

Such a strip was wound around a central core under clean conditions and the coil element was inserted into a device similar to that shown in Figs. 2-4.

The device was weighed dry and then rinsed with "Millipore" grade water for several hours to remove any residual cross-linking agent, after which the device was reweighed. The difference in mass indicated that the fill volume of the device was approximately 27 ml (i.e., the volume of liquid that the dry device can hold).

Procion Blue MXR dye was then coupled to the polymer layer by circulating through the device 100 ml of a solution containing 1 g dye, 30 ml 4 M NaCl, 02 ml 10M NaOH at 40ºC for 2 hours at a flow rate of 10 ml/min. The device was then rinsed with water for 2 hours. It was then rinsed with 600 ml 0.2M phosphate/1M NaCl solution, then again with water and finally with 0.14M PBS.

Albumin was separated from rabbit serum by circulating 50 ml of freshly thawed serum through the device at 20 ml/min for 2 hours, occasionally stopping the flow to allow good contact between the serum and the polymer layer.

The device was then rinsed with PBS until the wash solutions showed zero UV absorbance at 280 nm.

60 ml of 0.005M Tris/HCl/0.2M NaSCN was flowed through the device at a rate of 10 ml/min to elute 6 mg of albumin, followed by an additional 20 ml circulating through the device at a rate of 10 ml/min for 1/2 hour to elute an additional 6 mg of albumin.

Protein recovery was determined by optical absorption at 280 nm and electrophoresis was used to identify the recovered albumin.

Example 2 1. Synthesis of poly(2-hydroxyethyl methacrylate-co-methyl methacrylate-co-methacrylic acid-co-3-chloro-2-hydroxypropyl methacrylate) (16:1:1:2).

A one-liter three-neck round-bottom flask equipped with a condenser and a nitrogen inlet was charged with the following ingredients:

2-Hydroxyethyl methacrylate (0.48 moles) 62.45 g

Methyl methacrylate (0.03 moles) 3.00 g

Methacrylic acid (0.03 moles) 2.55 g

3-Chloro-2-hydroxypropyl methacrylate (0.06 moles) 10.72 g

p-Toluenesulphonic acid monohydrate 2.10 g

Bis(4-tert.butylcyclohexyl)peroxydicarbonate 0.79 g

Ethanol/Methylcellosolve (9:1 v/v) 250 ml

The solution was stirred at 50°C for 17 hours. During this period, nitrogen was bubbled through the solution. The polymer was recovered by precipitation in an excess of diethyl ether and dried in a desiccator. (Yield = 74.9 g).

2. Coating of the polymer

A coating solution was prepared consisting of 10% w/w of the above polymer in 100% dimethylformamide plus 10% w/w tetrabutylammonium hydroxide/polymer. 50 µm silica coated styrene beads were introduced in suspension into the solution as spacer beads.

The solution was coated onto one side of a corona discharged polyethylene terephthalate sheet to a wet coating thickness of 100 µm. The coating was dried at 90ºC for about 20 minutes. The other side of the sheet was coated in a similar manner except that the coating solution did not contain spacer beads. contained.

The coating was stable in water, saline solutions, ethanol, acetone and dimethylformamide, indicating that effective cross-linking occurred.

The coated sheet was cut into strips approximately 15 m long and 35 mm wide. One strip was wound around a central core under clean conditions and the coil element was placed in a device similar to that shown in Figures 2-4 and made of polypropylene. After placing the coiled element in the body of the device, the cap was applied by ultrasonic welding.

The device was rinsed with "Millipore" quality water.

3. Chromatographic use

Cibacron Blue F3G-A dye was coupled to the spooled element using the following procedure. 2.4 g of dye dissolved in 65 ml of water and 10 ml of sodium chloride solution (4M) was pumped through the previously rinsed device at a rate of 60 ml/min for 30 minutes. The device was left at room temperature for 90 minutes to allow the dye to be adsorbed onto the polymer surface. 1.25 ml of sodium hydroxide solution (10M) was added to the dye solution, which was then circulated through the device for 30 minutes at a pH of >12. The inlet and outlet parts of the device were sealed and the device was stored in a shaker at 25ºC for 2 days.

Water was flowed through the device at a rate of 60 ml/min until no dye could be detected by spectral absorption at 280 nm. Any uncoupled dye was removed from the coating surface by rinsing with a solution of sodium chloride (1M)/ethyl alcohol (25%).

The device was rinsed with 200 ml of a 0.05M phosphate buffer solution (pH 7) at a rate of 50 ml/min. 50 ml of untreated rabbit serum was introduced into the device at a rate of 30 ml/min. The first 30 ml of solution leaving the device was discarded and the remaining solution was circulated through the device for 30 minutes at the same flow rate. The pump was then turned off and the device was left for an additional 30 minutes. 0.05M phosphate buffer solution was passed through the device at a rate of 60 ml/min until the UV absorbance at 280 nm of the washing solutions was 0.

Elution was performed with a selected solution at a rate of 30 ml/min and the first six fractions were collected. The solution was circulated at a rate of 30 ml/min for 10 minutes and a seventh fraction was collected. The flow rate was increased to 60 ml/min and another 2x20 ml fractions were collected.

Using 0.2M NaSCN/0.05M Tris/HCl (pH 8) as eluent for a number of experiments, 3-11 mg albumin was obtained.

Protein recovery was determined by optical absorption at 280 nm and electrophoresis was used to identify the recovered protein.

Albumin recovery was also demonstrated using Procion MX-R dye instead of Cibacron F3G-A according to the procedure of this example.

Example 3

A polymer coated polyethylene terephthalate sheet was prepared as described in Parts 1 and 2 of Example 2, except that 100 µm silica coated styrene beads were used instead of 50 µm beads.

Samples of the coated polymer were activated by treatment with 4% divinyl sulphone solution in 0.5M sodium bicarbonate, pH 11. Rat anti-mouse K chain monoclonal antibody at a concentration of 0.8 mg/ml was coupled to the activated polymer coatings in a 0.1M sodium bicarbonate, 0.5M sodium chloride solution at pH 8.

The rat anti-mouse K chain antibody was purified from an ascites fluid obtained from Sera-lab (clone OX-20, code MAS 202C).

Jurkat cells, a human T-cell leukemia (a. Experimental Medicine 152: 1709, 1980; Gillis, S., and Watson, J.) grown in RPMI 1640 medium supplemented with 5% fetal calf serum (both from Flow Laboratories) were washed free of medium and resuspended in saline solution buffered with phosphate (PBS composition: 0.15M NaCl, 2.7 mM KCl, 8 mM Na₂HPO₄, 1.5 mM KH₂PO₄, pH 7.2). Cell viability was greater than 85% as judged by trypan blue exclusion.

A total of 3.1x10⁷ Jurkat cells (5.2x10⁷ cells/ml) were labeled with a mouse monoclonal antibody to the T cell surface antigen CD2. The antibody was obtained from Becton Dickinson Ltd. (Anti-Leu-Sb, catalog no. 7590). The antibody was added to the cells at a ratio of 0.5 µg antibody/10⁷ cells. The cells and antibody were incubated together, excess antibody was removed by centrifugation and the labeled cells were resuspended in PBS.

The polymer coatings to which OX-20 was coupled were incubated with the suspension of the labeled cells. The coatings were then washed with PBS and examined microscopically. The coatings showed a good uniform coating of bound cells at a high density.

For comparison purposes, samples of the coating were incubated with cells that did not show antibody labeling.

A subsequent microscopic examination showed that practically no cells had been bound to the polymer.

Claims (13)

1. Element for the selective removal of a chemical species from a fluid with an impermeable carrier sheet (10), the surface of which is designed in such a way that it is able to receive a ligand for the species to be removed, characterized in that a layer of a polymer (11) is adhered to the carrier sheet (10), which has functional groups on its surface through which the ligand is directly or indirectly covalently bound to the polymer. 2. Element according to claim 1, in which the thickness of the carrier sheet (10) is 0.05 to 0.5 mm. 3. An element according to claim 1 or claim 2, in which the polymer layer (11) is substantially hydrophilic. 4. An element according to any preceding claim, in which the polymer layer (11) is substantially non-porous. 5. An element according to any preceding claim, in which the polymer (11) is derived from an ethylenically unsaturated monomer having hydroxy groups. 6. An element according to any preceding claim, in which the polymer (11) is a cross-linked hydroxyethyl methacrylate. 7. Element according to one of the preceding claims, in which the thickness of the polymer layer (11) is 5 to 100 µm. 8. A method for selectively removing a chemical species from a fluid, which comprises contacting the fluid with an element according to any one of claims 1 to 7. 9. A device for the selective removal of a chemical species from a fluid comprising a housing (21) forming a chamber, the housing having a fluid inlet and a fluid outlet (23, 24), the chamber having at least one element (27) having a ligand for the species to be removed, the element or elements (27) being arranged relative to the inlet (23) and the outlet (24) in such a way that a flow path is formed such that, when the device is in use, fluid entering the chamber through the inlet (23) is passed over the surface of the element or elements (27) before the fluid leaves the chamber again through the outlet (24), characterized in that the element (27) is an element (27) according to any one of claims 1 to 7. 10. Device according to claim 9, with the element (27) in the form of a winding in which the turns of the winding are separated from one another by spacers (12) and in which the defined current path runs axially relative to the axis of the winding. 11. Device according to claim 9, with the element (27) in the form of a winding in which the turns of the winding are separated from one another by spacers (12), and the defined current path runs circumferentially through the windings. 12. Device according to claim 10 or 11, in which the spacing distance caused by the spacers (12) is practically constant. 13. Device according to claim 12, in which the spacing distance caused by the spacers (12) is 20 to 500 µm.